The coupling together of various information processing apparatuses by way of computer networks (hereinafter, referred to simply as networks) has resulted in much larger, more complex computer systems in recent years, and in line with this, a failure occurring in any of the information processing apparatuses has come to affect various information processing apparatuses on the network. As root cause analysis technique for identifying the locations and causes of these failures there is event correlation technique, which utilizes event information denoting the symptom of a failure notified from the information processing apparatus to analyze the failure location and cause (For example, Patent Literature 1). Because event correlation technique uses correlation with the event information notified at the time of the failure to infer the root cause, this technique has been used in diagnosing failures in network system for a long time.
Further, in Non-Patent Literature 2, there is disclosed a technique, which uses an expert system-based inference engine to rapidly determine the root cause of a failure by making rules for pairing up root causes that are inferred from a combination of the event information by this event correlation technique and events at the time of a failure.